Estimating the genetic burden: What really is deleterious in the human genome?

Tom Markello
Tom Markello
Speaker Name: 
Thomas Markello, M.D., Ph.D.
Speaker Title: 
Medical Geneticist, Clinical Staff Physician Office of the Clinical Director
Speaker Organization: 
National Human Genome Research Institute
Start Time: 
Wednesday, January 8, 2020 - 4:00pm
End Time: 
Wednesday, January 8, 2020 - 5:00pm
UC Santa Cruz Genomics Institute, Double Helix Conference Room B279
UCSC Genome Browser

Background: The problem of validating a candidate gene variant as the actual cause of a disease in a single human with a pathologic phenotype is the single most resource-intensive part of using a human genome sequence in medical diagnosis. Historical tests for validity in model animal and plant genetics required intentional breeding experiments. These are unethical in humans. Proxy validations have had lead to many spurious results in the history of agnostic sequencing. Finding “multiple independent alleles” in a natural population survey is the current gold standard when coupled with laboratory cell biology that explains a plausible expression mechanism.


Results: Recently we reported another perspective on answering this question using a cohort of patients with different rare highly penetrant diseases, matched to currently unaffected biological siblings in a fair and agnostic automated process to generate a very small candidate list of genetic burden variants for both individuals summed over each group. The assertion is that if the proband lists are generated identically as the siblings lists, and yet the sum of the candidates for the group of probands statistically exceeds the sibling’s summed list, then there is a high probability that the cause of many of the proband’s phenotypes are within the proband’s list. If the individual probands lists are small enough, this may assist in decisions for committing laboratory resources, like postdoctoral manpower, to biological validation work. The original project on exome sequencing has been extended to 50 cases done with whole genome sequencing and will be presented in this talk.

Event Type: